Maybe FPS drops?

No, it's warpy on most resolutions. 800x600 too for instance.

Sounds weird, how long has it been happening, and do you remember anything that changed that might have caused it, like new graphics drivers?

No, I haven't changed any drivers. After I switched from my CRT(to 1080p native) I assumed my lagometer had always been like that but it wasn't able to show me these tiny fluctuations due to low res.

Are you talking about just a 1 pixel difference? If yes, it's probably just because of rounding.

Yeah, 1 pixel, maybe 2 @ 1080p. But can you elaborate? Rounding in what, the display or connection code?

Rounding in the position on the screen to draw the lines. I haven't exactly looked at the lagometer drawing code in the Q3 source, but it's probably a floating point number floored to an integer which will probably be different for different screen heights.

If your ping is the same then don't worry about it, it's just the way it draws it.

Can you please look at the code weird posted and tell me what it means? I checked and the effect exists in 1440x900 and another res which was somethingX450, but I'm still wary because when I switch from 125 maxpackets to 100 in said resolutions the graph shows the same type of fluctuations, which to my mind speaks that it CAN show me the flaws.

There's just nothing to worry about. The lagometer just shows your pings, it doesn't change anything. Just because it looks uglier at a certain resolution doesn't mean your ping is any different.

hmmm if it's anywhere it should prolly be here:

if ( cgs.glconfig.vidWidth * 480 > cgs.glconfig.vidHeight * 640 )
{
*x += 0.5 * ( cgs.glconfig.vidWidth - ( cgs.glconfig.vidHeight * 640 / 480 ) );
}

in CG_AdjustFrom640()

doing tha math for X:
0.5 * ( 1600 - (900 * 640 / 480)) = 200.

0.5 * ( 1920 - (1080* 640 / 480)) = 240.

Seems like there is a difference.....

Afterwards there gets a lot of stuff also done though so I'm not 100% sure if this is it.
But it's the most clear strange thing I come across in the q3 code related to drawing the lagometer.

Maybe it's actually usefull to do this though... I'm not quite sure what it does hehe :D

If btw floats are casted into ints anywhere in the code then it can also be the screenXScale and (even more so the screenYScale)

after looking at it for a few secs while being fully awake...
The above code only seems to influence the width...

hmmm... It's more probable due to the vscale = range / MAX_LAGOMETER_PING;

Where the max ping = 900 and the range = 1/2*ah = 48 * cgs.screenYScale.

Where cgs.screenYScale = cgs.glconfig.vidHeight / 480.0;

Experiment:
on 1600 x 900 you have
screenYScale = 900/480 = 1.875.
1/2*ah = 1/2*48*1.875 = 45.

So ping X has a height of X*0.05px.


on 1920 x 1080 you have
screenYScale = 1080 / 480 = 2.25.
1/2*ah = 1/2*48*2.25 = 54.

So ping X has a height of X*0.06px.


Assumption that it actually gets rounded rather than floored ->
on height 900:
ping < 10 = 0 px
30 > ping > 10 = 1px.
50 > ping > 30 = 2px.


on height 1080:
ping < 8= 0 px
24 > ping > 8 = 1px
40 > ping > 24 = 2px.

I'd say that if you ping around 8 or around 24 you can see irregularities when your ping moves slightly on 1080 which you don't see on 900.

What's the next ping, cause I ping around 44?

well "around 44" probably means that jumps below 40 and sometimes above 40.

So on 1600 x 900 you'd have a stable line of 2 pixels. (as it's between 30 and 50 at all times).

On 1920 x 1080 you'd have an unstable line of 3 or 2 pixels. (3 if your ping > 40 and 2 if it's below 40)

I think that on 1920x1080 "per 16 ping" there's 1 pixel added. (starting from 8)
And on 1600x900 "per 20 ping" theres 1 pixel added. (starting from 10).

btw another thing I noticed once while I was playing on a japanese server (ye I'm weird like that).
Is that I'm not sure if the MAX_LAGOMETER_PING is actually still on 900 (as it is in the q3 code).
I thought that around ping 500 my lagometer was filled.... could just be a wrong/drunken memory though.

But as long as your "fluctuations" in the lagometer are just 1 pixel in size there is a high chance that the above is the reason.

Maybe the values are off though (if they changed MAX_LAGOMETER_PING).

if you're talking about green spikes in the lagometer, generally the larger your resolution gets, the larger and more spread out the spikes appear, at least in my experience with this monitor.

granted, i cant go higher than 1280x1024.

In 1600x900(non native) it's just a flat line - as flat as they get.

I noticed that on ALL my computers since the begining of QL, it is doing this because of the pixel scaling the lagometer image. But overall i would say higher resolution = more precise.

the higher resolution - the more precise it is... It's the same, just looks different

Like I said, it happens on lower resolutions(800x600) as well as higher(1080p).

do you realise that due to scaling between 1 pixel wide line and 2 pixels wide line there's a 'wavy' 1.5 or whatever pixel line?